Heriot-Watt University
ROS: The Research Output Service. Heriot-Watt University EdinburghNot a member yet
4689 research outputs found
Sort by
Integrated optimization of scale inhibitor squeeze treatment
Get PDFScale inhibitor (SI) squeeze treatment is one of the most widely adopted techniques to
control scale deposition. In this technique, chemical scale inhibitor is injected into the
near-wellbore area where it retains in the formation and then slowly releases in the
produced water when the well is back in production, preventing scale formation at a
concentration of few ppm.
The injection process normally starts with a preflush to condition the rock, and then the
main slug (containing the SI) is injected followed by an overflush to push the chemical
further deep into the reservoir. Before production, a shut-in period is also considered for
more SI retention in the formation. The aim of chemical inhibition is to delay the
deposition kinetics so that scaling issues are deferred from subsurface to surface, where
a much easier access allows easier handling of the deposition risk. The main goal of this
thesis is to present an integrated study of how to optimize the squeeze treatment design
based on the well conditions and by considering the operational constraints.
SI concentration, main treatment volume and the overflush volume are considered for
squeeze design optimization. Using the sensitivity study, the optimum inhibitor
concentration in the main slug is identified. The sensitivity results show that the most
efficient squeeze treatment is achieved when the SI is deployed with the highest possible
concentration, given the formation damage issues are avoided. In most cases, the well is
normally planned to be protected for a target lifetime, this will result in protecting the
well until the next treatment becomes available.
The squeeze lifetime function was shown to be differentiable against the squeeze
parameters, hence a gradient-based optimization algorithm, specifically Gradient Descent
(GD) algorithm was applied to optimize the main treatment and the overflush volume for
a given target squeeze lifetime. This will result in identifying the squeeze “Iso-Lifetime”
curve, which presents all the possible squeeze designs that provide the target lifetime,
using the optimum SI concentration. Based on the iso-lifetime designs, a cost analysis
was carried out to find the optimum treatment, where the CPB (total cost of squeeze per
barrel of water protected during the production period) was minimized, and the design
with the lowest CPB was selected as the optimum one.
For the cases with some flexibility in treatment lifetime, the same approach as described
above was employed for a range of target lifetimes to identify the optimum target. The
target lifetime that demonstrates the minimum CPB was identified as the optimum target
lifetime which can be considered to optimize the treatment in a single well for long-term.
Using this procedure, the optimum long-term strategy for squeeze treatment in a case
study was provided.
Multi-well squeeze design optimization was also studied in this thesis. Multi-well cases
may include scenarios such as treating two or more wells connected to a subsea manifold
or treating several single wells in the field where several wells of the same field are treated
simultaneously in a squeeze campaign. A supply vessel is normally used to deliver the SI
to the wells in a single trip. Due to the limitation of storage capacity on the vessel, the
amount of inhibitor which can be used is limited, hence the available amount of inhibitor
onboard should be optimally distributed among the wells. The squeeze campaign design
was optimized for two field cases, minimizing the total inhibitor volume and the total
downtime/pumping time, using the Multi-Objective Particle Swarm Optimization
(MOPSO) method. Once the wells are squeezed, they should all reach the target lifetime
of the campaign. This is essential such that all wells are protected until the next campaign.
Finally, the Pareto Front was identified for the field, including the optimum squeeze
campaign designs with the minimum cost, leading to the optimum inhibitor allocation
strategy.
The associated uncertainties with squeeze optimization were also considered in
optimization. These uncertainties are mainly related to the retention isotherm which is
normally derived by history matching. There might be several isotherms resulting in a
reasonable history match. This causes uncertainty in squeeze lifetime prediction.
Uncertainty quantification is considered in this research by evaluating the P10/P50/P90
percentiles using the likelihood function.
Finally, scale treatment optimization in geothermal reservoirs was investigated and the
most efficient scale treatment strategy in a geothermal doublet system was identified by
considering three different techniques of SI deployment: continuous injection downhole,
squeeze treatment and batch injection in the injector well. Optimum design for each of
the methods was studied considering different reservoir conditions, and the optimization
results were compared, providing the best scale treatment strategy in the reservoir
Relational knowledge and representation for reinforcement learning
Get PDFIn reinforcement learning, an agent interacts with the environment, learns from feedback about the quality of its actions, and improves its behaviour or policy in order
to maximise its expected utility. Learning efficiently in large scale problems is a major challenge. State aggregation is possible in problems with a first-order structure,
allowing the agent to learn in an abstraction of the original problem which is of
considerably smaller scale. One approach is to learn the Q-values of actions which
are approximated by a relational function approximator. This is the basis for relational reinforcement learning (RRL). We abstract the state with first-order features
which consist of only variables, thereby aggregating similar states from all problems
of the same domain to abstract states. We study the limitations of RRL due to
this abstraction and introduce the concepts of consistent abstraction, subsumption
of problems, and abstract-equivalent problems. We propose three methods to overcome the limitations, extending the types of problems our RRL method can solve.
Next, to further improve the learning efficiency, we propose to learn different types
of generalised knowledge. The policy is influenced by directed exploration based on
multiple types of intrinsic rewards and avoids previously encountered dead ends. In
addition, we incorporate model-based techniques to provide better quality estimates
of the Q-values. Transfer learning is possible by directly leveraging the generalised
knowledge to accelerate learning in a new problem. Lastly, we introduce a new class
of problems which considers dynamic objects and time-bounded goals. We discuss
the complications these bring to RRL and present some solutions. We also propose a framework for multi-agent coordination to achieve joint goals represented by
time-bounded goals by decomposing a multi-agent problem into single-agent problems. We evaluate our work empirically in six domains to demonstrate its efficacy
in solving large scale problems and transfer learning
Meta-interpretive learning of proof strategies
Get PDFIn modern mathematics, mechanised theorem proving software is playing an ever increasing role. By enlisting the help of computers mathematicians are able to
formally prove more complex results than they perhaps otherwise could, however
those computers are still incapable of drawing many of the conclusions which would
be obvious to a human user and so human intervention is still required.
In this thesis we consider the use of an adapted machine learning technique to
begin addressing this issue. We consider the use of proof strategies to provide a
high-level view of how a proof is structured, including information about why a
particular step was taken. We extend the Metagol meta-interpretive learning tool
to facilitate learning these strategies. We begin with a small set of examples and
refine our approach, demonstrating the improvements experimentally. We go on to
discuss the learning of more complicated strategies, some of the issues faced in doing
so and how we could address them. We conclude by evaluating the experiments as
a whole, identifying the weak points in our approach and suggesting ways in which
they can be addressed in future work
Benthic ecosystem functioning of the western Clarion-Clipperton Zone, Pacific Ocean, and the West Antarctic Peninsula : a study to assess the effectiveness of Areas of Particular Environmental Interest (APEIs) in the context of deep-sea mining and the effects of climate change
Get PDFThe deep sea encompasses the largest ecosystem on Earth and remains largely unexplored.
With plans for deep-sea mining and the increasing impacts of climate change on our oceans,
there is a growing necessity to understand and safeguard deep-sea biodiversity and ecosystem
functioning. The cycling of carbon (C) by deep-sea benthic communities is a key ecosystem
function and pulse-chase experiments are aimed to measure this process I conducted pulse-chase experiments in situ at abyssal depths (4800-5300 m) in three no-mining areas, called
Areas of Particular Environmental Interest (APEIs), in the Clarion-Clipperton Zone (CCZ) and
in ex situ experiments using sediments collected from a bathyal (500-600 m) fjord, Andvord
Bay, and the continental shelf of the West Antarctic Peninsula (WAP). My results underline
the importance of organic C in driving ecosystem dynamics at the abyssal seafloor and support
the notion that Antarctic fjords are hotspots of benthic biomass and ecosystem functions. The
microbial community was shown to be a key player in the short term (1.5 d) cycling of C on
the abyssal plain of the western equatorial Pacific Ocean, which is consistent with other
published studies, while the macrofaunal community (>300 µm) dominated the initial (~1 d)
degradation of phytodetritus in Andvord Bay. My study provides important information on
benthic ecosystem functioning in the western CCZ, an area targeted for commercial-scale deep-sea mining, and the WAP, a region that is becoming increasingly impacted by climate change
Exploring determinants of self-service technology success in German food retail : a retail technology manufacturer case
Get PDFThis study explores strategic success determinants of self-service technologies (SSTs)
in the German retail food industry of quick- and self-service restaurants from the
perspective of a retail technology manufacturer. The current state of the academic
literature primarily focuses on explaining influencing factors on the customer adoption
rate of SSTs by consulting information systems success models and related technology
acceptance theories. The underlying theoretical frameworks are based on DeLone and
McLean’s updated Information Systems Success Model and on the third version of the
Technology Acceptance Model. Variables related to customer experience and
satisfaction are predominantly put into the focus of research projects available. Little
attention though is being paid to non-customer-oriented success dimensions covering
information quality or system quality of the SSTs under observation. Moreover,
businesses developing and providing SSTs to retailers seem to be disregarded in the
existing literature as well.
As a single-case study design, this research programme seeks out to explore strategic
SST success determinants via insights gathered from a major retail technology
manufacturer delivering SST solutions to food retailers in Germany. Based on the data
collected in semi-structured interviews conducted with industry experts from senior and
top management functions in the company, strategic SST success factors are identified
and aggregated into an overarching model of SST success for the retail food industry of
quick- and self-service restaurants in Germany. The core focus thereby is on selfordering kiosks and self-checkout solutions as the most commonly used types of SSTs
in this market. A pilot study was executed prior to the main study and demonstrated the
methodology and research strategy to be appropriate for the research project, which
bases its conceptual framework on the updated DeLone and McLean’s IS Success
Model and concepts found in the third version of the Technology Acceptance Model.
This research project contributes to the academic literature by providing a detailed
collection of SST success determinants and dimensions relevant in the German retail
food industry of quick- and self-service restaurants, which are arranged in a newly
developed SST Success Model for this concrete use case in German food retail. The
results of this study are of value for retail practitioners adopting self-service strategies
and implementing SST solutions in store environments of quick- and self-service
restaurants
Novel, comic-based approach to smartphone permission requests
Get PDFThis doctoral thesis investigates comics as a medium for presenting permission requests
to users of smartphones and assessing how they can be used to support the users in making
more informed decisions. Through three empirical studies, this thesis has generated
insights into the creation of comic-based permission requests, the impact they have and
what users would expect from them. In the first study, three co-design workshops were
run where participants created their own comic requests, generating a series of design
considerations for comic-based permission requests to better inform users. These
considerations were used to design and develop the comic requests used in the subsequent
studies to investigate if they supported users in making informed decisions in a balanced
way. The comic-based permissions were evaluated by users in an online survey (Study
2) and in a WebApp (Study 3), to affirm the viability and the reliability of the
considerations. The results of the evaluations suggest that comics are a viable medium
to display more informative permission requests which help inform users and promote
more informed decisions.James Watt Scholarshi
Historical context, preference, and capabilities : a case study of indigenous peoples’ homelessness experiences in Seattle, USA
Get PDFIn order to create effective responses that acknowledge the diversity of homelessness and
its contributing factors, it is beneficial to look at extreme cases. Indigenous peoples
continually rank near the bottom of nearly every US social, health, and economic
indicator (Alba et al. 2003). Furthermore, data consistently indicate that Indigenous
peoples remain overrepresented among the homeless population in the US.
This thesis aims to explain the experiences of homelessness amongst Indigenous peoples
in the US and the impact that current emergency accommodation options have on their
ability to live a well-lived life. It uses Martha Nussbaum's (2011) Capability Approach
– encompassing ten "essential" capabilities required to live a well-lived life – to inform
the analysis of the empirical case study data presented. The thesis begins by investigating
the extent to which US colonialism and historical housing and homelessness policies
account for the disproportionate risk of homelessness experienced by Indigenous peoples.
It moves on to examine the extent to which the currently available forms of emergency
accommodation impact on the capabilities of Indigenous peoples, as understood through
the lens of Nussbaum’s framework. Finally, Nussbaum's capabilities are further drawn
on while seeking to explain the role played by the characteristics of emergency
accommodation in accounting for the decision of some Indigenous peoples in the US to
sleep rough.
This thesis uses a case study approach with the city of Seattle as its focus, which has the
third-largest homeless population in the US, and a highly disproportionate number of
Indigenous peoples affected. The fieldwork undertook a total of 14 interviews with key
stakeholders and 30 interviews with Indigenous peoples experiencing homelessness in
Seattle, as well as a review of local statistics and research reports.
This thesis argues that centuries of colonialism, racially segregated housing, housing
discrimination, and forced relocation affected Indigenous peoples’ ability to accumulate
generational wealth and secure housing. Other findings suggest Indigenous peoples
experiencing homelessness often prioritise control over their safety and health while also
giving importance to cultural identity in navigating emergency accommodation options.
The environmental conditions of emergency shelters required Indigenous peoples to
compromise and prioritise between capabilities and often, as a result, influenced their
decision to sleep rough. Emergency shelter options in Seattle do not support physical
recovery, provide safety, or support the facilitation of connecting to other necessary immediate and long-term resources and housing programmes. The thesis also argues that
Nussbaum's Capability Approach should be modified by changing the definitions of
specific capabilities to accommodate Indigenous peoples’ experiences of homelessness
Approximations in actuarial and financial mathematics
Get PDFThis thesis consists of three topics that are related to approximations.
We first investigate the accuracy of Taylor polynomials in approximating utility
functions. We show that increasing the polynomial order does not necessarily improve the approximation of the expected utility. The proofs use methods from the
theory of parabolic second-order partial differential equations.
In the second part of the thesis, we aim to analyse the spread process in a short
time in the SIS epidemic model of computer networks. We show that the short-time asymptotics of infection probability depends on the network structure. We use
concepts and methods from graph theory and defined in this chapter.
In the third part of the thesis, we propose a single-network-based algorithm
using deep learning techniques where neural network is used to approximate the
derivatives of a function. We provide computational underpinnings for applying the
replacement closeout convention in the valuation of a defaultable financial claim
with counterparty credit risk.
Numerical examples illustrate all results in the three parts
Data augmentation to enhance human-robot interaction
Get PDFContemporary research in robotics is focusing on autonomous and independent
robots that can operate in unconstrained environments and interact with humans. Perceiving of and acting within these environments not only creates novel challenges
but provides new opportunities to improve state-of-the-art of robotics. This thesis
presents one such opportunity by utilising information from Human-Robot Interaction for data augmentation.
Data augmentation is a valuable tool to improve recognition approaches, especially
in complex environments. Recent research in machine learning is often driven by the
mantra "we need more data". In contrast, data augmentation can outperform state of-the-art methods provided additional information can be used without the need to
acquire more input data. In Human-Robot Interaction, we benefit from knowing
the context in which a robot acts and perceives information. The presented work
takes a closer look at such use-cases, introduces the notion of direct and indirect
data augmentation, and highlights the importance of data augmentation by providing
experiments on datasets that were created from interactions with the iCub humanoid
robot.
This thesis shows the application of data augmentation and provides improvements
for state-of-the-art action and object recognition in the context of Human-Robot Interaction
Development of kinetic-theory-based models accounting for charge transport in polydisperse gas-solid flows
Get PDFKinetic-theory-based transport models are developed for polydisperse granular and gas-solid flows with contact electrification. Starting with the Boltzmann-Enskog kinetic
equations, a transport equation for the solid phase charge is introduced into kinetic theory
for granular flows with, first, monodisperse particles and, latter, binary solid mixtures.
For binary mixture, each solid phase possesses its own macroscopic quantities (i.e. solid
volume fraction, mean velocity, granular temperature, mean charge and charge variance)
using the non-equipartitioning of random fluctuating kinetic energy. The primary model
is extended for dilute regime where self-diffusion of charge is modelled via a charge-velocity correlation. The model is further extended for granular flows far away from
equilibrium conditions by applying a perturbation to the Maxwellian state of particle
velocities. The hydrodynamic and the additional charge transport models are assessed
through hard-sphere simulation results at each stage of the model development. The charge
evolution is well predicted for granular flows at equilibrium conditions, while predictions
of the flows at non-equilibrium conditions are less accurate. The mathematical models are
implemented into an open-source multiphysics computational framework (OpenFOAM)
with developing a new solver. This solver is then used to study the effect of vessel
size on charge build-up in gas-solid suspensions and to model lightning during volcanic
eruptions
Access Repository Dashboard
Do you manage ROS: The Research Output Service. Heriot-Watt University Edinburgh ? Access insider analytics, issue reports and manage access to outputs from your repository in the CORE Repository Dashboard!